925 articles tagged with Benchmark
Nowy benchmark PerceptionBench ujawnia, że topowe modele AI nie radzą sobie z podstawowym widzeniem, osiągając wyniki poniżej 60% skuteczności. Rzekome błędy logiczne maszyn okazuj...
This week Anthropic launched Claude Opus 5 at half the price of its flagship, positioned it as...
AI referral traffic remains a small share of website visits, but it is becoming too important to...
`AI coding assistants have become part of everyday software development. They can generate...
Mi:dm K 2.5 Pro hits 80.9% on MMLU-Pro but only 10% on long-context reasoning — independently measured, not self-reported.https://olud.ai/leaderboard.html#LLM #Benchmarks #OpenSour...
Benchmark AI cho dev bằng cùng một prompt và cùng đoạn code để chọn model theo workload thật. Gemini3.7 Flash có785 điểm ...
Airbnb Eval-driven development: Lessons from evaluating GenAI at scaleArticle URL: https://medium.com/airbnb-engineering/eval-driven-development-lessons-from-evaluating-genai-at-sc...
Честный RAG eval set: как собрать первые 100-300 кейсов и не обмануть себя цифройВ прошлой статье серии мы сжимали эмбеддинги и замеряли, насколько изменится выдача относительно по...
Nearly every AI agent benchmark you read is unfalsifiable. Not wrong, necessarily - unfalsifiable....
Redwood Research et Anthropic lancent le Conceptual Reasoning Index, benchmark évaluant décision théorique, argumentation philosophique et cohérence logique sans vérité vérifiable....
By Nithin Alkanand, Partner, Sapphire Human Capital Capital has stopped being the binding constraint on India’s artificial intelligence ambition. Capability has taken its place. Sp...
MindTopo reveals VLMs’ spatial reasoning abilities. Microsoft Research unveils MindTopo, a new benchmark exposing critical gaps in VLMs' spatial reasoning and planning for robotics...
SPIEval: Evaluating Large Language Models as Mobile Assistants over Scattered Personal Information. New benchmark SPIEval reveals critical gaps in LLM-based mobile assistants' abil...
1 Bloomberg: The #Kospi #benchmark has soared 22% in the past 10 sessions, putting it back in a #bullmarket. The Kospi trades like a pure play on #AI because two big winners from t...
タムズにもAnthropicがあったらよかったですねAgentic reliability and evaluations : Enterprises that got burned by a bad eval are the most likely to remove humans from the loop, not the least https://...
I recently ran a small China AI benchmark for eight luxury-jewelry brands. The most interesting...
Toward Human Rights Benchmarking for LLMs: A Pilot Methodology. First-of-its-kind benchmark HumRightsBench evaluates LLMs on human rights law reasoning, revealing significant perfo...
BEAM - the Benchmark for Evaluating Agent Memory - is a good benchmark because it tests memory the...
A few weeks ago I wrote about the six questions I ask every new model. The response I got most often...
A new model drops roughly every other week now, and my timeline fills up with the same two claims:...
Everyone is building AI memory systems. But how do we know which ones actually work? As AI agents...
Last month I wrote about the six questions I ask every new model. Those questions are good for a...
Grok 4.6 matches GPT-5.6 Sol's reasoning benchmark at one-fifth the output cost, per independent testing. For agents running many prompts before completion, the pricing gap compoun...
VibeLifeBench introduces a benchmark of 200 long-horizon tasks to test if LLM agents can act proactively and persistently in a changing simulated world. Seven leading models all sc...